Home
About SampleSmith
About the Demos
About the Tutorials
About the Soundfonts
Articles
Downloads
Midi Editing
Favourite Links
Contact
Purchase
Some Basic Guitar Picking Principles for the Keyboardist
Strumming Mechanics
Separating Ups from Downs
Placing the Strum
Strum Feel
Strum Feel Part II
Strumming Mechanics
Back To Top
Probably everyone has seen a guitarist strumming a fairly complex syncopated rhythm at one time or another, and maybe you've noticed how smoothly the right hand is pumping up and down, without much of a clue as to where the syncopation is coming from. Looking a little closer will reveal that on many of the up and down strokes you're seeing, the strings are not being played at all. That is, the guitarist is setting up the picking hand motion as a subdivision on the time, maybe 1/8's 1/16's or 1/8 triplets, but still holding back the decision to actually strike the string. At the same time he may be holding the shape of a particular chord with the left hand, but again hold the decision to press the strings into contact with the frets or just use the left hand to mute the strings.
So now that he has established the fundamental rhythmic subdivision, he can choose at will to use combinations of up and down strokes together with muted notes, ringing notes, or silence to create any complexity of rhythm based on that subdivision.
Here are a couple of examples in 4/4 time:
Fig.1
The first example shows, on the top staff a fairly common syncopated rhythm part notated in the usual way. The lower staff shows the constant down/up motion of the pick through the same rhythmic pattern. The guitarist simply misses the strings with the pick on the tied notes.
Fig.2
The second example shows a very common rhythmic strum pattern in 2 variations. In both cases the picking hand is again just pumping up and down, and the left hand is lifted just enough to deaden the strings immediately after the down stroke on beat 2, then replaced in time for the following upstroke on the 'and' of 2. On the lower staff we can see that the guitarist has opted to miss striking the strings on the down stroke of beats 3 and 1 to give a more syncopated feel.
Separating Ups From Downs
Back To TopStrumming Mechanics
Separating Ups from Downs
Placing the Strum
Strum Feel
Strum Feel Part II
This is pretty simple stuff, and absolutely basic to guitar playing. My main reason for explaining it is to arrive at a method for extracting the ups from the downs in our midi file, and this is because they sound different, and ideally will need a different EQ treatment if we are going to use the same samples for both pick directions. Having determined that the upstrokes are going to fall on the 'and' of 1/8 note strums, or on the second and fourth stroke of a 1/16 based figure, we can use a logical editor to extract these upstrokes and put them on channels separate from the down strokes.
Now, in order to achieve this separate EQ on the two different channel sets, we may use a soundfont player like the
RGC SFZ+
(multiple outputs) where we could separate the upstrokes to different channels and assign them to a different output with its own set of EQ settings. Almost as effective, once we've separated to different channels, would be to use midi Brightness controls on those channels, or failing that we might make a copy of the midi file and mute the down strokes in one file and the upstrokes in another, and run 2 instances of
SFZ
(single stereo output) with different track settings on each instance, taking care to keep the pan and gain settings consistent.
In the preparation of the SampleSmith demos. I used a single instance of SFZ, and employed a combination of Velocity, Brightness and Attack Time to differentiate the strokes.
Why should the upstrokes and downstrokes sound different? There are a few good reasons for this. The first and most obvious is gravity. Even if the guitarist wants to make the 2 strokes sound identical, he or she has gravity working with them on the down stroke and against them on the upstroke and this will inevitably make a difference in the sound. The second reason is that with the motion of the arm and wrist, the pick's angle of attack will usually be a bit different on the 2 strokes. And the 3rd reason is really a question. If you could make them sound different, then why, as a player, wouldn't you use this device for expressive purposes? You can change the pick angle on the upstrokes, or pick closer to the bridge on either the up or down, and so on.
Placing the Strum
Back To TopStrumming Mechanics
Separating Ups from Downs
Placing the Strum
Strum Feel
Strum Feel Part II
Another thing about strumming that comes so naturally it's easy to overlook, is whether the strum should begin or end on the target beat. At a decent tempo, we're talking about fractions of a second here, so it's difficult to illustrate in notation or tablature.
First let's look at a slow strum, perhaps near the end of a tune, to get the idea of what I'm talking about.
Fig.3
Here the strum is played in such a way that the final note lands on beat 1, while in the next example,
Fig.4
the strum is begun on beat 1 with a totally different musical effect. This would be especially evident if we had a drummer and bassist coming down on beat one in both cases.
Now this may seem an obvious distinction, and it certainly is, in the slowly strummed examples I've given. What about the case of an up tempo bluegrass rhythm part, or a chord melody in a jazz solo? If you've just laid out a 'boom-ching' sort of rhythm part and it's quantized, which way do you go when you start to insert the 'inter-string delays' we did in the strum demo editing example? You could move the lower strings forward a few ticks, or the higher strings back like we did in Home Sweet Home, and which way you go will have a big effect on the feel of the piece.
Strum Feel
Back To TopStrumming Mechanics
Separating Ups from Downs
Placing the Strum
Strum Feel
Strum Feel Part II
Here, we are talking about feel, and I can only offer the fruits of my own experience, and hope I can explain it well enough. To illustrate, let's look at 3 midi files. I'll call them 'Quantized', 'Laid Back' and 'On Top'. The second and third are both valid examples of what one might play in two different tunes at the same tempo, but with two different feels. The 'Laid Back' is a lazy, rolling kind of feel, with the guitar feeling like it's just behind the beat -not slowing down at all, it's just a tiny bit on the late side all the time. The 'On Top' version is a more driving feel, with the strum on the back beat being really on top -even leading the beat a bit.
The guitaristic problem here is that when playing a full chord with a pick, one can only approximate an instant in time. The strum, no matter how quickly done, has duration, and if you don't get that duration in your midi file, it won't sound like a strummed guitar. Certainly one can play 4 or even 5 note chords using the fingers of the right hand to the same effect that a piano can get by picking them all at (almost) the same time, (and this is very useful in classical or latin-american music) but it doesn't have the same feel as a bluegrass rhythm player shoving out a driving backbeat with a heavy pick, or a Django style chunker.
With these 3 files, I started with 8 bars of a simple boom-ching bass note and strum on the back beats. All six strings were placed on separate channels. Bass notes were quantized on beats 1 & 3, and the back beat strums quantized on beats 2 & 4, and all velocities are at 100.
I copied this 'Orig Quant' file to 'Laid Back' where I delayed channels 3, 2 & 1 (3rd, 2nd and first strings) by 15, 21 and 24 ticks respectively.
Fig.5
I also copied the 'Orig Quant' file to 'On Top' where I advanced channels 4, 3 & 2 ( 4th, 3rd and 2nd strings) by 25, 9 and 4 ticks respectively.
Fig.6
You'll be able to see in the screenshots, that in the case of the 'Laid Back' file the strums begin on the beat, and the 'On Top' strums end on the beat, increasing the sense of drive in the track.
Use the links, below to access audio examples of this.
Strum Feel - Quantized 250bpm.mp3
(Machine gun midi.)
Strum Feel - Laid Back 250bpm.mp3
Strum Feel - On Top 250bpm.mp3
Strum Feel - Quantized 120bpm.mp3
Strum Feel - Laid Back 120bpm.mp3
Strum Feel - On Top 120bpm.mp3
And the Midi File:
Strum Feel.mid
for import into your sequencer.
I've added a simple bass drum and side-stick to give a firm grasp of where the time really is in all three files, and you'll notice that in the quantized files, the guitar is instantly recognizable as midi, with the strums showing no time delay between strings, at all. It's only provided as a reference. When you play the 'Laid Back' and On Top' these files, you'll notice a difference in the feel. It's fairly subtle. This is nowhere near the extremes you could use, but I hope it demonstrates the necessity of avoiding rigidly quantized strums. Even if you are emulating a classical or latin (fingerstyle) technique, you'lll still find that some variation in 'note on' timing is very important to getting a human feel into the track.
I found it very interesting to note that when the 2 files were played back with the Guitar soloed,(all files have identical velocities), the same sound font and player, and no EQ or other processing on either channel, the 'On Top' played back 0.5dB louder then the 'Laid Back' file. I'm assuming that would be because the bass note had less time to decay before the strum was added to it in the case of the 'On Top file. Regardless of the reason, simply changing the timing has increased the physical energy of the track, which would further contribute to the feeling of drive.
Strum Feel Part II
Back To TopStrumming Mechanics
Separating Ups from Downs
Placing the Strum
Strum Feel
Strum Feel Part II
As I mention again in the article on editing Home Sweet Home, these Inter-string delays are going to be personal to each player, and to how he or she feels the drive of the tune or interprets the lyric content, and that will of course be affected by the tempo. Having said that, it would be nice to have some starting numbers.
We might make a basic rule that the extremes for a given tempo and feel, let's say 100 bpm for a fairly bouncy tune, for now, are 21 ticks. Of course me might expand that to something longer for a more mellow, romantic or introspective kind of feel, and something a bit shorter for a brighter, happier feel.
Fortunately, working in ticks is automatically tempo-compensating.
21 ticks at 100 bpm =60/100*21/480 = 0.026sec
At 200 bpm 21 ticks =60/200*21/480 = 0.013sec
So, even though it's nice to know in milliseconds what the delays might need to be, to project a certain feel, it's entirely possible, and a lot more practical to eliminate the actual time in ms from the equation and just think in ticks. This assumes that psychologically the delays work if they are proportional to tempo. I can't answer that one for sure, but it seems reasonable, at least within a sensible range of tempi, so that's what I'm working on for now.
Here's a starting point:
Fig.7
>
All of this is subject to so many variables and preferences that these numbers can only be regarded as arbitrary points of departure. As you gain experience, you'll come up with some rules of thumb that work for you. I offer the above as a suggestion for cataloguing your own findings. First of all you'll want to experiment with making the upstroke delays longer or shorter than the down strokes, and this will add more character or 'human-ness' to the piece.
Now having established the basic delay values, we have the option as mentioned at the beginning of this section of leading the beat, following it, or even spreading the delays out so the beat falls somewhere in the middle of the arpeggio (which is really what a strum is -played very quickly). And, again, the choice will depend upon the aesthetic values of the music at the time. We might use all of these variables in one tune, at some time or other.
One other aspect that influences these delays is the flexibility of the individual strings. The original delay values I came up with for my own use were derived from some experiments I did with strums at different tempos. The upstroke delays between the first 2 strings (E and B) were always shorter than the down stroke delays between wound strings (D to G), or the 3rd to 2nd string. While much of this must have to do with the acceleration of the hand on the down stroke, that explanation doesn't seem to follow for the upstroke where the hand is beginning its acceleration indicating a delay that would be greater instead of smaller.
When beginning any strum, and particularly on the upstroke to the first string, the pick is guided through space by muscle memory alone. It's not bouncing off a previous string, and given the geometry changes of the hand and arm through the stopping of the previous down stroke and setting up the pick angle for this upstroke, the chances are pretty good that the E string is going to see more 'pick' than say, the D string on a down stroke. And given that the unwound strings are more flexible than the wound strings, when that E string gets a lot of pick, it bends more, and tends to 'bunch up with the B string a bit -maybe more than just a bit if the player is feeling aggressive. So then the B and E strings may tend to be released at very close to the same time.
This would have the effect of increasing the delay before the E string is released and reducing the delay between the E and B strings. It would also have the effect of making the initial attack of the E string a bit sharp- a bit of the 'boing' effect you get when over stressing a string before releasing it.
All this minutiae may seem like overkill, but you can bet these are the things I'd be thinking about if I were working on a guitar part for an important soundtrack. I hope this stuff helps to get you thinking about some of the simple physics that go into guitar playing and guitar imitating. It all boils down to first principles eventually.
Back To TopStrumming Mechanics
Separating Ups from Downs
Placing the Strum
Strum Feel
Strum Feel Part II
Last updated Feb 01/06 19:00 MST
© SampleSmith.com